The large, spherical part in the picture below might look like a daunting machining project, but for Keller Technology, this sort of work is common—high-value, custom jobs that typically involve low production quantities, complex geometries, no room for error and no second chances. For jobs like this, proving out machining processes before tools meet metal is critical, although that doesn’t diminish the importance of timely deliveries. Steve Ziff, CAD/CAM manufacturing engineer, says one particular system is critical to meeting both of these potentially conflicting goals: Vericut simulation software from CGTech (Irvine, California). Since installation a few years ago, the software has eliminated the need for time-consuming manual prove-outs, improved confidence on the shop floor and enabled faster setups.
Keller Technology specializes in equipment used to manufacture semiconductors; medical devices; food products; commercial and industrial products; and equipment for research at universities and national labs. The company traces its roots to 1918 as a general job shop servicing customers in metropolitan Buffalo, New York. The next 40 years saw the company evolve to provide machine-building services to customers throughout the United States. Today, the fifth-generation family business serves customers around the world. In addition to its manufacturing facility near Buffalo, it maintains an assembly plant in Charlotte, North Carolina as well as affiliates in Seoul, South Korea.
The spherical part mentioned above is a component for a high-resolution diffractometer used by the Department of Energy. It started as a forged piece of aluminum measuring a
little more than 1 meter in diameter and weighing more than 3,000 pounds. Machined on an SL 100 five-axis machine from Parpas America (Bloomfield Hills, Michigan), the part required removing more than a ton of material to bring it to its final weight to 610 pounds. Like much of Keller Technology’s work, the part was a one-off requiring high precision, custom tooling and relatively lengthy machining routines, and it had to be done right the first time—welding to fix any gouges was not an option.
Prior to implementing Vericut, the company likely would have proved out the part program manually by running the machine through a mock cycle on the shop floor, Mr. Ziff says. That alone would have wasted a great deal of valuable machining time, not to mention the hours shopfloor personnel would have spent properly orienting the part. Cutter clearance was a particular concern because the forged aluminum sphere and its custom fixture barely fit within the machine’s workzone. This problem was exacerbated by the fact that the 51 square, tapered pockets located throughout the part surface measure as deep as 14 inches in some cases, necessitating relatively lengthy cutting tools.
Still, a manual prove-out would have been more reliable than the most readily available alternative: the simulation capability embedded in the company’s CAM system. The
problem, Mr. Ziff ex-plains, is that the shop’s CAM-integrated verification capability is limited to checking only the internal CAM file. As a result, the system evaluates only the cutting tool’s interaction with the part without accounting for other factors that affect the process. “In the real world, you’ve got a lot more things to worry about—doors, clamps, angle plates, how the part fits on the table,” Mr. Ziff explains. “There was no representation of the full machining environment.”
Vericut provides just such a representation, with user-defined virtual models of all equipment involved in the production process, Mr. Ziff says. Upon installation, Keller programmers input information for every machine on the shop floor, including not only axis travels and other physical dimensions, but also feeds, speeds, horsepower and other attributes. Access to that sort of information enables the system to issue a warning in the event of any discrepancies between the program and the actual equipment capabilities, such as when a CAM file calls for 10,000 rpm on a machine capable of only 3,000.
Custom tooling, fixtures and part models are typically passed into the simulation session though an interface with the CAM system. The resulting virtual environment provides
the context for simulations that run from the same post-processed G code used by machine tool controls. As opposed to CAM-integrated systems that are limited to only the internal CAM file, this ensures a more comprehensive rendering of how the process will proceed on the shop floor. “What Vericut does differently is that it puts the code itself in the driver’s seat,” Mr. Ziff says.
To explain, he cites the example of a tool change. The simulation module in the company’s CAM system would depict the machining routine proceeding seamlessly with the new tool instead of the old. That’s perfectly fine for determining the viability of the tool path itself. What’s lost, however, are the spindle’s moves away from the workpiece, toward the machine’s carousel for the tool swap and back again to the point of the cut, all while avoiding any potential collisions. Simulation based on the actual G code accounts for such moves, which are particularly important for parts like the diffractometer component that might cause clearance issues.
Comprehensive simulation within a realistic rendering of the machining environment saved significant time and money during the planning phases of that job, Mr. Ziff says. CAM-integrated verification would have had little to offer in terms of designing the part’s custom fixture, orienting it within the workzone and determining the proper lengths for the custom tools used to machine it. Manually performing all the trial and error required for this process would have been both time and cost prohibitive. With Vericut, however, the labor involved amounted to only mouse clicks. “It’s easy for us here in programming to, say, move the fixture 10 inches and see what happens,” Mr. Ziff explains. “Doing that out on the machine would take a very long time.”
While difficult even by Keller Technology’s standards, the diffractometer component job provides an example of how Vericut provides the confidence needed to move forward with machining high-value parts without the time associated with manual prove-outs. “With some large forgings approaching a quarter million dollars, we don’t have the luxury of just throwing another piece of stock on the machine after a mistake,” Mr. Ziff says.
Programming a robot with the same CAM software used for a multi-axis machine tool makes it unnecessary to “teach” the robot by jogging it manually from point to point and recording these point-to-point moves as the robot’s motion commands. Robotmaster is a software package that provides this CAD/CAM-based, off-line programming for robots. It runs fully integrated inside Mastercam CAM software for CNC machine tools.
The additional rotary milling axis on these machines allows them to complete many types of complex parts in a single setup, but these machines have gained a reputation for being difficult to program. Today’s CAM software, however, eases the programming challenge significantly.
The many rules of programming CNC cutter radius compensation can make machining problems difficult to diagnose and correct.